high end interfaces and linux

[Klangfarben]

Originally Posted by BethHarmon

But I would still argue not a different worth parting with $3500. Once you have your final release with plugin effects mixed and mastered at 44.1/48k and 16-bit dithered (because this is more than enough for any human ears) played on end user systems the differences will indeed be subtle to the point of not being able to pass a blind AB test.

Again, very much disagree here. Most film scores are being recorded at 24/96 and 24/192. Some might call that overkill but the music also is being passed on elsewhere after the music mix is done to the dub, dolby encoding, etc. and the higher quality you deliver it, the better quality you get at the end of the chain.

Also be very careful with your 16-bit 44.1 is "more than enough" argument. It was a compromise when the redbook standard was made and still is a compromise today. Which is why the industry has shifted to 24/48 and 24/96 for most music recording. Are you going to be able to tell the difference on your laptop speakers or a bluetooth speaker? Of course not. But on really good speakers and good monitoring equipment it's a different story.

And just on a personal note, I'm kind of disappointed when someone on this sub-forum asks what are some high end interface options on Linux and most everyone chimes in with "you don't need a high end interface". I think it would be better if people genuinely tried to give OP some options.

[chmaha]

Originally Posted by Klangfarben

Again, very much disagree here. Most film scores are being recorded at 24/96 and 24/192. Some might call that overkill but the music also is being passed on elsewhere after the music mix is done to the dub, dolby encoding, etc. and the higher quality you deliver it, the better quality you get at the end of the chain.

Also be very careful with your 16-bit 44.1 is "more than enough" argument. It was a compromise when the redbook standard was made and still is a compromise today. Which is why the industry has shifted to 24/48 and 24/96 for most music recording. Are you going to be able to tell the difference on your laptop speakers or a bluetooth speaker? Of course not. But on really good speakers and good monitoring equipment it's a different story.

And just on a personal note, I'm kind of disappointed when someone on this sub-forum asks what are some high end interface options on Linux and most everyone chimes in with "you don't need a high end interface". I think it would be better if people genuinely tried to give OP some options.

This is a lot of faulty logic. 16-bit contains all the dynamic range we humans could ever need (difference between a jack hammer and a gnat and being able to hear both in the same room would lead to hearing loss) and 44.1k contains more than the very best hypothetical human hearing limits of 20kHz.

Passing on audio (and recording, editing and mixing) at 24-bit or 32-bit float is fantastic practice (due to the amount of headroom) but the end product needs only ever to be 16-bit dithered. This is the sole purpose of 24-bit recording: headroom for capture, editing and mixing. There is also no need to capture analogue audio at 96k and 192k given all sound captured at 44.1/48k can be perfectly recreated analogue to digital and vice versa. I'll accept virtual synths can generate very high frequency content that can be folded back at lower sample rates but that's not what we are talking about here. Additionally you must have excellent hearing to appreciate 96k/24-bit recordings. Spoiler alert: you can't for the aforementioned limits of human hearing and that inconvenient brownian noise that affects every piece of electronics on the planet and completely covers 24-bit dither. In reality, equipment is only capable of about 20-bits of dynamic range but, again, completely unusable unless you really have a serious death wish for your ears. Yes, 16-bit 44.1k really is enough for humans now and forever. Even classical music doesn't come close to using even half the available dynamic range of 16-bit.

And, to your final point, I feel it's my duty to provide a balanced view informed by science as to why a supposed "high-end" interface is definitely not a requirement and driven in large part by marketing BS (384k, 32-bit integer etc).

[Largos]

Originally Posted by Klangfarben

Also have to agree to disagree here. The quality of the converter makes a big difference. Just because most people won't notice a difference doesn't mean it's at the same level.

So, which adc are good and which are bad?

I presume you are here to give the OP some info and not just flex on your special ears.

[Venn]

Picked up the RME AIO Pro last month and it falls into what I consider the high end.

The hdspmixer app from 2003 still works and hdspeconf is able to adjust I/O levels, clocksource, AES and the like.

It was (basically) a drop in replacement for the 9632 in the studio.

Not having TotalMix on Linux was a bummer but everything runs through Reaper during the live streams.

[audiojunkie]

Originally Posted by Klangfarben

And just on a personal note, I'm kind of disappointed when someone on this sub-forum asks what are some high end interface options on Linux and most everyone chimes in with "you don't need a high end interface". I think it would be better if people genuinely tried to give OP some options.

For my part, I honestly don't know of any class compliant audio cards that come with DSPs that Linux can access. Most audio cards for Apple systems are class compliant, and would thus likely be usable on Linux, but again--the DSP part comes into play. I would love to learn of some of those devices being made, just so that I could recommend them. I recognize that everyone has different needs, and professionals use the very best. But alas, I'm a hobbyist who hasn't played in a band for over 20 years. I try to help where I can, when I can.

Do YOU know of any audio cards that have on board DSP that Linux can use?

[Klangfarben]

Originally Posted by Largos

So, which adc are good and which are bad?

I presume you are here to give the OP some info and not just flex on your special ears.

Er, I guess you missed the part where I recommended a high end interface directly after the part you quoted. Also, not sure what in the world you considered "flexing" but given the fact you didn't seem to read the rest of my post with the actual information/recommendation part I'm not going to overly worry about it.

One option you could try is the Merging Anubis. I have one and both the converters and preamps are excellent. One of the best boxes I've purchased (and I've purchased a lot). The only real limitation is it only has 4 inputs and 2 mic pres. But if you can get away with 4 in/6 out then it might be a good option for you on Linux.

There are drivers from Merging here.

https://github.com/dewiweb/ALSA-RAVENNA-AES67-Driver

And someone else forked the Merging driver and also added a WebUI for controlling the settings which is better than Merging's release.

https://github.com/bondagit/aes67-linux-daemon

[Venn]

Originally Posted by audiojunkie

For my part, I honestly don't know of any class compliant audio cards that come with DSPs that Linux can access. Most audio cards for Apple systems are class compliant, and would thus likely be usable on Linux, but again--the DSP part comes into play. I would love to learn of some of those devices being made, just so that I could recommend them. I recognize that everyone has different needs, and professionals use the very best. But alas, I'm a hobbyist who hasn't played in a band for over 20 years. I try to help where I can, when I can.

Do YOU know of any audio cards that have on board DSP that Linux can use?

MOTU 828ES: Web GUI

MOTU 828MK3(firewire): ALSA with snd-firewire-ctl-services

[audiojunkie]

Edit: Deleted for redundancy and to "stay out of it".

[Klangfarben]

Originally Posted by audiojunkie

Do YOU know of any audio cards that have on board DSP that Linux can use?

The Merging Anubis can direct monitor the inputs as well. It runs pretty impressive software right on the unit itself which you can control from the touchscreen. One of the things you can control is switching the monitoring between the live inputs and computer inputs.

[audiojunkie]

Originally Posted by Venn

MOTU 828ES: Web GUI

MOTU 828MK3(firewire): ALSA with snd-firewire-ctl-services

That's definitely a nice looking piece of hardware. However, I don't see a confirmation of either Class Compliancy or direct Linux support. I do however notice that the device has some included iPad software for controlling the routing. Has it been confirmed that this device (and its DSP) will work with Linux? If so, this is quite amazing!

[audiojunkie]

Originally Posted by Klangfarben

The Merging Anubis can direct monitor the inputs as well. It runs pretty impressive software right on the unit itself which you can control from the touchscreen. One of the things you can control is switching the monitoring between the live inputs and computer inputs.

That is admittedly a VERY nice piece of hardware. I had never heard of it before today. Can you confirm that the built-in effects can be accessed while using it with Linux?

I like to help wherever I can, but I'm still learning like everyone else. We need people (like you) to speak up when you know of cool equipment like this. My intentions were right, and I certainly would have mentioned this had I known about it. Thanks for posting!

[audiojunkie]

Originally Posted by BethHarmon

Nice-looking perhaps but the Behringer UMC series manages 110 db(A) weighted dynamic range. More than enough for humans (and the same chips used by Apogee apparently).

That's great! So there is something nice for both ends of the price spectrum. Having used Behringer over the years and knowing the reputation of Uli Behringer, I haven't had a great opinion of the company (although the synth recreations that have been coming out have done a lot to help with the reputation). I couldn't pay anyone to take my Behringer equipment 20 years ago, so I hope the quality has truly improved.

Edit: I just looked over it again, and I don't see that the Behringer has the required linux accessible onboard DSP. The Behringer may not meet the needs that were being requested.

[chmaha]

Originally Posted by audiojunkie

That's great! So there is something nice for both ends of the price spectrum. Having used Behringer over the years and knowing the reputation of Uli Behringer, I haven't had a great opinion of the company (although the synth recreations that have been coming out have done a lot to help with the reputation). I couldn't pay anyone to take my Behringer equipment 20 years ago, so I hope the quality has truly improved.

I simply use the Behringer model as a typical mid-to-low budget USB interface. You could add Focusrite, Presonus, MAudio, SSL, Audient etc to the list. All have excellent chips inside that don't cost the earth.

Anyway, Klangfarben is going to conveniently ignore the science I provided earlier so there's not much point me continuing to contribute to this particular thread.

[Venn]

Originally Posted by audiojunkie

That's definitely a nice looking piece of hardware. However, I don't see a confirmation of either Class Compliancy or direct Linux support. I do however notice that the device has some included iPad software for controlling the routing. Has it been confirmed that this device (and its DSP) will work with Linux? If so, this is quite amazing!

I had the 828ES over a weekend to play with. Long as you are running a 5.x kernel and roll the firmware back to V1 it justworks™.

Honestly, most of the MOTU AVB series appear work as class compliant devices but people are still reporting issues with channel jumping. I'll pick one up to play with when I get some time.

The MOTU MK3 series have been supported by FFADO for a long time. However, Tak has them working with his ALSA stack snd-firewire-improve.

Last year he added the ability to control the DSP function on MK3 devices using snd-firewire-ctl-services from alsamixer.

I can confirm that works on the 828MK3 and the Traveler MK3.

[audiojunkie]

Originally Posted by BethHarmon

I simply use the Behringer model as a typical mid-to-low budget USB interface. You could add Focusrite, Presonus, MAudio, SSL, Audient etc to the list. All have excellent chips inside that don't cost the earth.

Anyway, Klangfarben is going to conveniently ignore the science I provided earlier so there's not much point me continuing to contribute to this particular thread.

True, but for Norbury's requested use, there needs to be a linux accessible built in DSP. I don't believe any of these provide what he was needing.

[audiojunkie]

Originally Posted by Venn

I had the 828ES over a weekend to play with. Long as you are running a 5.x kernel and roll the firmware back to V1 it justworks™.

Honestly, most of the MOTU AVB series appear work as class compliant devices but people are still reporting issues with channel jumping. I'll pick one up to play with when I get some time.

The MOTU MK3 series have been supported by FFADO for a long time. However, Tak has them working with his ALSA stack snd-firewire-improve.

Last year he added the ability to control the DSP function on MK3 devices using snd-firewire-ctl-services from alsamixer.

I can confirm that works on the 828MK3 and the Traveler MK3.

Very cool! The next time I need to look for another audio interface, I'll have to consider this series.

[chmaha]

Originally Posted by audiojunkie

I just looked over it again, and I don't see that the Behringer has the required linux accessible onboard DSP. The Behringer may not meet the needs that were being requested.

Huh? I'm not sure what you are talking about. The onboard DSP is a physical "direct monitor" button on the 202HD and a "Mix" physical knob between direct in and playback on the higher models. There are no DSP effects such as compressor or reverb.

My argument is that onboard DSP is a bit of a red herring when low-latency software monitoring with plugins is so accessible.

[Glennbo]

Originally Posted by audiojunkie

That's great! So there is something nice for both ends of the price spectrum. Having used Behringer over the years and knowing the reputation of Uli Behringer, I haven't had a great opinion of the company (although the synth recreations that have been coming out have done a lot to help with the reputation). I couldn't pay anyone to take my Behringer equipment 20 years ago, so I hope the quality has truly improved.

The older Behringer units with Xenyx preamps were okay, but not in the same league as the newer Behringer units with Midas pres.

Before getting my UMC1820, I was using a Behringer MX1602 with Xenyx pres with a pair of M-Audio Delta 2496 cards. I could hear a distinct improvement switching to the UMC1820, and most of it I attribute to the Midas mic pres, because the M-Audio Delta series had very decent converters.

[Klangfarben]

Originally Posted by audiojunkie

That is admittedly a VERY nice piece of hardware. I had never heard of it before today. Can you confirm that the built-in effects can be accessed while using it with Linux?

Yes. The dsp is controlled on the interface itself via its touchscreen and then can be routed wherever you want (it routes via ethernet) including the computer. As far as plugins, there isn't much right now. Merging makes a de-esser and you can also run speaker calibration software on it. As well as the direct monitoring. They also opened up the plugin dsp to third parties but I don't think anyone besides Merging has developed anything yet. The antelope interfaces have a lot more plugin options. I don't really care about the plugin part, just the direct monitoring part. For 2 grand it's a ridiculously impressive box.

[audiojunkie]

Originally Posted by BethHarmon

Huh? I'm not sure what you are talking about. The onboard DSP is a physical "direct monitor" button on the 202HD and a "Mix" physical knob between direct in and playback on the higher models. There are no DSP effects such as compressor or reverb.

My argument is that onboard DSP is a bit of a red herring when low-latency software monitoring with plugins is so accessible.

Yeah, I don't think Norbury wants "low-latency" monitoring. He's looking for no-latency monitoring with onboard DSP effects. Otherwise, I'd be totally recommending any device that has a physical "direct monitor" button--because it is perfectly sufficient for me. But I'm a hobbyist and Norbury is a professional with professional clients and requirements. That's why I didn't recommend or mention knowing of any Linux compatible audio devices with onboard DSP.

[audiojunkie]

Originally Posted by Glennbo

The older Behringer units with Xenyx preamps were okay, but not in the same league as the newer Behringer units with Midas pres.

Before getting my UMC1820, I was using a Behringer MX1602 with Xenyx pres with a pair of M-Audio Delta 2496 cards. I could hear a distinct improvement switching to the UMC1820, and most of it I attribute to the Midas mic pres, because the M-Audio Delta series had very decent converters.

Yes, I think I still have my M-Audio Delta 4x4 somewhere in a box. I liked those devices too.

[Glennbo]

Originally Posted by BethHarmon

Huh? I'm not sure what you are talking about. The onboard DSP is a physical "direct monitor" button on the 202HD and a "Mix" physical knob between direct in and playback on the higher models. There are no DSP effects such as compressor or reverb.

The ones with the mix knob, like the UMC1820 will let you mix direct dry input signal, with post DAW signal that has FX, but that still only works at low latency settings in REAPER.

I just leave my mix knob all the way to the right, post REAPER and monitor with FX coming from REAPER. Never once have I wanted to crank the knob to the left and monitor direct. Same goes for musician friends of mine who have recorded here in my studio.

[audiojunkie]

Originally Posted by Klangfarben

Yes. The dsp is controlled on the interface itself via its touchscreen and then can be routed wherever you want (it routes via ethernet) including the computer. As far as plugins, there isn't much right now. Merging makes a de-esser and you can also run speaker calibration software on it. As well as the direct monitoring. They also opened up the plugin dsp to third parties but I don't think anyone besides Merging has developed anything yet. The antelope interfaces have a lot more plugin options. I don't really care about the plugin part, just the direct monitoring part. For 2 grand it's a ridiculously impressive box.

It certainly sounds like it! I could have sworn that I read that it not only had the de-esser but also a compressor and reverbs. I can't seem to find it now, but then again, I've been looking at so many audio adapters today.

[Glennbo]

Originally Posted by audiojunkie

Yes, I think I still have my M-Audio Delta 4x4 somewhere in a box. I liked those devices too.

The two Delta 2496 cards I used effectively made a Delta 44, because the two units could both record and playback 4 in and 4 out, with one card supplying the clock to the other over S/PDIF.

[Klangfarben]

Originally Posted by BethHarmon

Anyway, Klangfarben is going to conveniently ignore the science I provided earlier so there's not much point me continuing to contribute to this particular thread.

Sorry to try and keep the thread from devolving into yet another pointless "you can't even hear it" thread and try to actually provide OP with information to help his request.

If you don't think I understand the science (or the math) you would be very mistaken. You have your opinion based on professional experience and I have mine based on the same. Let's just leave it at that and stop the pointless attacking.

[audiojunkie]

Originally Posted by Glennbo

The two Delta 2496 cards I used effectively made a Delta 44, because the two units could both record and playback 4 in and 4 out, with one card supplying the clock to the other over S/PDIF.

I seem to have forgotten that part. I think the only time I've seen mine in the past 20 years was when I was moving to my current place. I'm pretty sure that I saw it in one of my many totes of stuff that I was moving. That was probably about 5 years ago. It was my first "professional" audio card I ever owned, and I did all of the recording in my band with it. The recording quality sucks, but I can only blame myself at that time, and not the hardware.

[EcBaPr]

i think you can use a Behringer X32 with full functionality on linux.. it has CC drivers to communicate with the DAW and all the DSP/routing is controlled through network.. last i looked their mixer GUI software runs on linux so (I think) you get 100% of the functionality ?

people may scoff at a digital mixer for studio use but in my tests X32 latency seems about the same as RME UFX.. the X32 pre amps sound great and the internal routing/FX options blow most studio interfaces away. the X32 rack maintains the form factor, you dont need the console..

i dont currently use linux but cant you aggregate interfaces with jack or pipewire ? in other words use an interface you think has nice pre amps (and class compliant drivers) as the tracking inputs and a digital mixer like X32 for playback/monitoring ?

you could record X32 FX if you wanted, they sound pretty good but if the goal is mainly monitoring FX while tracking, then X32 is well qualified for that.. not to mention loads of routing options to each musician which helps etc..

[Glennbo]

Originally Posted by audiojunkie

I seem to have forgotten that part. I think the only time I've seen mine in the past 20 years was when I was moving to my current place. I'm pretty sure that I saw it in one of my many totes of stuff that I was moving. That was probably about 5 years ago. It was my first "professional" audio card I ever owned, and I did all of the recording in my band with it. The recording quality sucks, but I can only blame myself at that time, and not the hardware.

I bought two M-Audio Delta 2496 cards in 2000. One in a dedicated GigaStudio96 machine for drum samples and the other for the DAW on a separate machine. I kept expecting Windows to obsolete the M-Audio cards, but even old drivers kept working. Then I switched to Linux and the cards still functioned perfectly.

It was because I knew I would be building a new DAW computer and it would definitely not have PCI card slots that I bought a UMC1820.

[cporro]

Originally Posted by BethHarmon

My argument is that there are no obvious quality differences between $500 and $4000. No offense meant by that. I think it's the audio world equivalent of edible gold leaf used by fancy restaurants If world-renowned classical engineers feel comfortable recording a whole orchestra using just the preamps on a MixPre II, I think we don't have to worry too much.

that's it. take a shoot out with your ears and find out. would love to see the shoot out between this "high end" gear and the cheap stuff. this is a different time. great audio quality can be had for cheap. people should be worried about how long they can productively use their stuff. how easy will it be to use?

this idea of expensive gear producing emotionally better (by way of fidelity) music is a pitch. from a guy who went through the moultonlabs golden ears series 2 times and scored better then most even down to 1/3 octaves. i went down that silly rabbit hole for year. waste of time. especially today. maybe we should all be humbled by A/B testing software. try it. you'll be surprised by how well you don't hear.

[cporro]

Originally Posted by BethHarmon

This is a lot of faulty logic. 16-bit contains all the dynamic range we humans could ever need (difference between a jack hammer and a gnat and being able to hear both in the same room would lead to hearing loss) and 44.1k contains more than the very best hypothetical human hearing limits of 20kHz.

Passing on audio (and recording, editing and mixing) at 24-bit or 32-bit float is fantastic practice (due to the amount of headroom) but the end product needs only ever to be 16-bit dithered. This is the sole purpose of 24-bit recording: headroom for capture, editing and mixing. There is also no need to capture analogue audio at 96k and 192k given all sound captured at 44.1/48k can be perfectly recreated analogue to digital and vice versa. I'll accept virtual synths can generate very high frequency content that can be folded back at lower sample rates but that's not what we are talking about here. Additionally you must have excellent hearing to appreciate 96k/24-bit recordings. Spoiler alert: you can't for the aforementioned limits of human hearing and that inconvenient brownian noise that affects every piece of electronics on the planet and completely covers 24-bit dither. In reality, equipment is only capable of about 20-bits of dynamic range but, again, completely unusable unless you really have a serious death wish for your ears. Yes, 16-bit 44.1k really is enough for humans now and forever. Even classical music doesn't come close to using even half the available dynamic range of 16-bit.

And, to your final point, I feel it's my duty to provide a balanced view informed by science as to why a supposed "high-end" interface is definitely not a requirement and driven in large part by marketing BS (384k, 32-bit integer etc).

whut she said. you need to stop reading. stop believing. start listening.

test how much at 20hz you hear. approx nyquist upper limit of 44.1khz. my hearing drops off sharply at 19.5k. and all the info is centered around 1k. truth is... good hearing at 16k is fine for making music. sheesh.

from one person on the interwebs to another. i don't care about being right. ego gone. but i will set the BS straight. it's in everyone's interest.

[Largos]

Originally Posted by Klangfarben

Er, I guess you missed the part where I recommended a high end interface directly after the part you quoted. Also, not sure what in the world you considered "flexing" but given the fact you didn't seem to read the rest of my post with the actual information/recommendation part I'm not going to overly worry about it.

I didn't ask you about interfaces, I asked you about adc, the one specific component that you have decided to single out.

[Mcgiver69]

Originally Posted by norbury brook

actually that's not how it works you can always hear what you've recorded but the monitoring of your 'new' track is zero latency.

I've dropped singers in on a nearly finished mix to do a repair ,where by buffers are at 1024 samples because there's so many plugins and they have a zero latency monitor mix with DSP reverb.

ASIO Direct monitoring NOT available on OSX it's the reason I chose windows when setting up a Native DAW back on the day.



M

Which in that case you can bounce your project to a track, open another tab, load the track and record your singer at a lower sample rate while keeping the project you're mixing open in another tab. Once you're done with the singe just copy and paste or drag and drop the vocal track into your existing project. Voila!!!!

[norbury brook]

Originally Posted by Mcgiver69

Which in that case you can bounce your project to a track, open another tab, load the track and record your singer at a lower sample rate while keeping the project you're mixing open in another tab. Once you're done with the singe just copy and paste or drag and drop the vocal track into your existing project. Voila!!!!

yes, i can see that working the only issue is you're fixed with the mix at that point. usually people when tracking will want things changing on the fly, ''can you take the drums down/Up, can you lose the BV's they're putting me off, can I have more bass, turn the guitar down in the last verse'' etc etc

None of this is possible with a fixed 'bounce'


M

[chmaha]

Originally Posted by norbury brook

yes, i can see that working the only issue is you're fixed with the mix at that point. usually people when tracking will want things changing on the fly, ''can you take the drums down/Up, can you lose the BV's they're putting me off, can I have more bass, turn the guitar down in the last verse'' etc etc

None of this is possible with a fixed 'bounce'

You can always bounce stems. But, yes, at this point, it's getting overly complicated

[norbury brook]

Originally Posted by BethHarmon

You can always bounce stems. But, yes, at this point, it's getting overly complicated

Actually, thats a good idea. By mix time i have everything going to 'Mix busses' so if i bounced the 5 or 6 mix busses as stems then that would solve most of those issues.

in conjunction with the web remote that could give 'the turn' an adjustable cue mix on their phone



Excellent. I'm going to try this.


M

[bjohn]

Originally Posted by cporro

test how much at 20hz you hear. approx nyquist upper limit of 44.1khz. my hearing drops off sharply at 19.5k. and all the info is centered around 1k. truth is... good hearing at 16k is fine for making music. sheesh.

I don't think the people who advocate higher sampling rates believe a) that a higher number of samples per given unit of time produce more accurate waveforms (there is no evidence for that) or b) that there's a benefit to capturing frequencies beyond the range of human hearing.

I think it's more about the desire to push any distortion effects from the low-pass antialiasing filter in the analog stage of the A/D converter above the range of hearing so they don't affect the higher audible frequencies. That filter is applied at half the sampling rate, and at 44.1 any distortion caused by the filter could creep into the audible range.

There's a short discussion of that here: https://www.sonarworks.com/blog/lear...ng-sample-rate

I'd guess this is the most likely underlying cause (apart from expectation bias) for people claiming that 48k, 98k, or higher sampling rates sound better to their ears.

Another reason to use higher sampling rates is purely commercial: there's a market for "high definition" audio and some people are willing to pay a premium for it regardless of whether that makes any objective sense.

Tony Faulkner, the renowned classical music engineer who uses a MixPre 10t to record orchestras for the BBC (for broadcasts only, at 48/24) records for CD and other formats using a different system at high sample rates to "future proof" his recordings to meet potential future demand from audiophiles. Then there's the DSD market. High-definition PCM or DSD may not make sense from an objective perspective, but people do claim to hear differences and thus they are willing to pay for high-definition audio. If you want to sell to that market, that will dictate your equipment and your sampling rates.

[norbury brook]

the bottom line is to use your ears. If you can hear a difference then go for it.

I could definitely hear a difference between my old UR824 and the new AXR4 , so end of story it stayed . I do this with everything I get for the studio.

Also, remember that the actual AD/DA chips are only half the picture, there's a whole analogue stage before and after that makes a huge difference. Also things like the power supply make a difference.

I've found from experience that 'cheaper' audio interfaces sound much better @96k , the difference is noticeable, whereas at the high end the difference between 44.1 and 96 is less due to the better quality analogue side of the interface which is what you're paying for really. AD/DA chips are cheap, it's the analogue and power supply design and parts that cost the money.

M

[karbomusic]

Originally Posted by norbury brook

Here' my Guitar amp wall in the studio

Do I see Two-Rocks there? I picked up a Studio Signature Combo a couple months back... OMFG is all I can say - I own and have played through tons of amps over the years, and OMFG, OMFG, OMFG. I'm sure you know what I mean.

[bjohn]

Originally Posted by norbury brook

Also, remember that the actual AD/DA chips are only half the picture, there's a whole analogue stage before and after that makes a huge difference. Also things like the power supply make a difference.

Yes--it's probably the analog stage that accounts for the differences people hear between A/D converters; most of them use the same chips but blind tests reveal clear differences.

Clocking is also a factor: I've gotten a small but noticeable improvement in sound on my MixPre 10t by clocking it externally with a Grimm CC2. I blind-tested it with a friend who could hear differences although she actually preferred the sound of the MixPre without the external clock whereas I preferred it with the external clock! To me the bottom end gained a noticeable increase in clarity and a feeling of "tightness" when externally clocked.

[chmaha]

Also worth bearing in mind is that even the budget USB interfaces internally upscale in terms of sample rate so recording higher to avoid distortion effects is probably not necessary.