Bit depth question

Hello,
My audio device can sample up to 24-bit. When I create a 32-bit file in Audition 2.0 and record material using the audio device, Audition tells me the file is 32-bit, and it is indeed twice the size of an equivalent 16-bit file. But is it really a 32-bit file, and could there be any issues with the file? It seems fine in every way.
Thank you.

No audio hardware actually samples at greater than 24-bit, because there's absolutely no point - even 24-bit depth isn't actually usable in full; a system like this could in theory digitise a noise floor way lower than can be physically achieved by any mic and preamp system available - you'd need at least liquid nitrogen cooling of all the components before you even started to look at the rest of the problems!
So why does Audition record in 32-bit? Well, 32-bit Floating Point digitising is a bit different. The actual 24-bit signal is recorded quite faithfully (although not quite in the form of an integer signal) and the other 8 bits are essentially left as zeros during recording. What they actually are is scaling bits. And this comes in seriously useful when processing. What it means is that your original signal can be scaled up and down without loss. In an integer engine, if you decided to throw away 30dB of a signal, saved the result and reopened the file and amplified it again, you'd find that your 24-bit signal was effectively 19-bit. In Audition, if you did exactly the same thing with a 32-bit Floating Point signal, you wouldn't lose any bit depth at all. No it's not magic - it's just the effect of storing the original data in a form that inherenently doesn't get modified when an amplitude change is asked for - it's only the scaling system that does, and this doesn't contain audio data.
So yes it's a real 32-bit signal - but not all of those 32 bits are used until you do some processing.

Similar Messages

Maximum audio sample rate and bit depth question

Anyone worked out what the maximum sample rates and bit depths AppleTV can output are?
I'm digitising some old LPs and while I suspect I can get away with 48kHz sample rate and 16 bit depth, I'm not sure about 96kHz sample rate or 24bit resolution.
If I import recordings as AIFFs or WAVs to iTunes it shows the recording parameters in iTunes, but my old Yamaha processor which accepts PCM doesn't show the source data values, though I know it can handle 96kHz 24bit from DVD audio.
It takes no more time recording at any available sample rates or bit depths, so I might as well maximise an album's recording quality for archiving to DVD/posterity as I only want to do each LP once!
If AppleTV downsamples however there wouldn't be much point streaming higher rates.
I wonder how many people out there stream uncompressed audio to AppleTV? With external drives which will hold several hundred uncompressed CD albums is there any good reason not to these days when you are playing back via your hi-fi? (I confess most of my music is in MP3 format just because i haven't got round to ripping again uncompressed for AppleTV).
No doubt there'll be a deluge of comments saying that recording LPs at high quality settings is a waste of time, but some of us still prefer the sound of vinyl over CD...
AC

I guess the answer to this question relies on someone having an external digital amp/decoder/processor that can display the source sample rate and bit depth during playback, together with some suitable 'demo' files.
AC

Final cut pro millions of colours + bit depth question

Hello
I am working in final cut pro 7 and I wanted to know what is the maximum bit depth I can export using the Prores codec? All I see in compression settings for rendering my timeline when wanting to render with Prores 4444 is the option for 'millions of colors' and 'millions of colors +' I was under the impression that millions of colors refered to 8 bit... does the alpha channel mean I can get 10 bit? can the alpha channel hold 2 more bits per channel or something? Or is there no way I can export a 10bit file using the Prores codec within fcp7..? is it all just 8bit. -and when I select 422HQ there is no advanced options for millions of colors..what does this mean? is the only way to get 10bit out of fcp7 to render with the 10bit uncompressed codec? and if so can I render the timeline in prores while im working with it then delete all the renders and change the render codec to 10bit uncompressed, will this now be properly giving me 10bit from the original 4444 12 bit files i imported in the beginning..?
Any help is much appreciated

ProRes is 10-bit. Every ProRes codec is 10-bit...LT, 422, HQ. Not one of them is 8-bit. Except for ProRes 444...that's 12 bit.

After Effects 8-bit depth question

Hey everyone, this is something i've just been wondering about. Correct me if i'm wrong but its my understanding that with 10-bit depth footage or higher After effects crunches it to 8-bit after the render/export. If that's the case then what is the actual point of capturing at higher bit depths in the first place if we can't maintain them when it comes down to final delivery?
Thanks very much for any help.

Cacogen24 wrote:
Hey everyone, this is something i've just been wondering about. Correct me if i'm wrong but its my understanding that with 10-bit depth footage or higher After effects crunches it to 8-bit after the render/export.
You're only partly wrong. If you keep the project 8bit then you're right, but that's a terrible workflow. Where most folks foul up is in the rendering.
After Effects offers 8bit, 16bit, and 32bit processing options. If you render a 32bit project to an 8 bit codec, you get 8bits, but if you render a 32bit project to a 10bit codec (Black Magic, ProRez, or several other flavors of higher bit depth codecs) you get a 10bit product. You can also render to 16 or 32 bit image sequences.
All of these settings are available in the Adobe Media Encoder or through the Output Module settings in the Render Cue. Since I work primarily on a Mac all of my production renders are to ProRez 422 or 444 (10bit) unless I'm freelancing at a facility that works with another capture card. Most places I work use Black Magic cards so we render to the Black Magic 10 bit codec.
My standard workflow is to start with an 8bit project, set up the project and render motion and compositing tests, then switch to a 32bit project settings, do the color correction, then render to a production codec that is at least 10bit.
BTW, don't use the export feature in AE for any kind of delivery. Export's only professionally usable export is Flash Professional XFL if you need animated motion or text in a flash project or Premiere Pro Project if you want to use Premiere to edit your AE project.
I hope this helps.

Maximum Bit Depth /Maximum Render Quality Questions

Maximum Bit Depth
If my project contains high-bit-depth assets generated by high-definition camcorders, I was told to select Maximum Bit Depth because Adobe Premiere Pro uses all the color information in these assets when processing effects or generating preview files. I'm capturing HDV using the Matrox RTX-2 Hardware in Matrox AVI format.
When I finally export my project using Adobe Media Encoder CS4, will selecting Maximum Bit Depth provide better color resolution once I post to Blu-ray format?
Maximum Render Quality
I was told that by using Maximum Render Quality, I maintain sharp detail when scaling from large formats to smaller formats, or from high-definition to standard-definition formats as well as maximizes the quality of motion in rendered clips and sequences. It also renders moving assets more sharply. It's my understanding that at maximum quality, rendering takes more time, and uses more RAM than at the default normal quality. I'm running Vista 64 Bit with 8 GIGs of RAM so I'm hoping to take advantage of this feature.
Will this also help to improve better resolution when I finally export my project using Adobe Media Encoder CS4 and post to Blu-ray format?
Does it look like I have the specs to handle Maximum Bit Depth and Maximum Render Quality when creating a new HDV project with the support of the Matrox RTX 2 Hardware capturing in Matrox AVI format? See Below Specs.
System Specs
Case: Coolmaster-830
Op System: Vista Ultima 64 Bit
Edit Suite: Adobe Creative Suite 4 Production Premium Line Upgrade
Adobe Premiere Pro CS 4.0.1 update before installing RT.X2 Card and 4.0 tools
Performed updates on all Adobe Production Premium Products as of 03/01/2009
Matrox RTX2 4.0 Tools
Main Display: Dell 3007 30"
DVI Monitor: Dell 2408WFP 24"
MB: ASUS P5E3 Deluxe/WiFi-AP LGA 775 Intel X38
Display Card: SAPPHIRE Radeon HD 4870 512MB GDDR5 Toxic ver.
PS: Corsair|CMPSU-1000HX 1000W
CPU: INTEL Quad Core Q9650 3G
MEM: 2Gx4|Corsair TW3X4G1333C9DHXR DDR3 (8 Gigs Total)
1 Sys Drive: Seagate Barracuda 7200.11 500GB 7200 RPM 32MB
Cache SATA 3.0Gb/s
2 Raid 0: Seagate Barracuda 7200.11 500GB 7200 RPM 32MB Cache SATA 3.0Gb/s Using Intel's integrared Raid Controller on MB

Just some details that i find useful on maximum render depth
You really need it even with 8bit source files, when using heavy grading/multiple curves/vignettes. If after grading you see banding, go to sequence > sequence settings from the top menu and check "maximum bit depth (ignore the performance popup), then check again your preview (it will change in a second) to see if banding is still present in 32bit mode. If no banding, you must check it when exporting, if banding is still there, change your grading, then uncheck it to continue with editing.
Unfortunately Maximum bit depth exporting is extremely time-consuming, but can really SAVE YOUR DAY when facing artifacts after heavy grading, by completely or almost completely eliminating banding and other unwanted color distortions.
Use it only for either small previews or the really final output.
Best Regards.

Reduce bit depth or convert color profile first? (best practices question)

For making final deliverable files from working files, is it best to convert to a new color profile before reducing bit depth? Or vise versa?
Our working files are 16 bit with the ProPhoto color space. Our deliverable files are 8 bit AdobeRGB tiffs and sRGB jpegs. We convert using relative colorimetric with black point compensation. Does it make a difference which order these changes are made in?
Thanks in advance for your help!

A profile conversion recalculates RGB values, so yes, it should be done in 16 bit depth.

Why only 10-bit depth dng files from 16-bit Nikon D90 nef files?

When I convert 16-bit .nef files from my Nikon D90 to DNG I get only 10-bits depth.
Since the camera should be producing 12-bit depth it seems I am losing information in the conversion, and I don't want that.
I have installed the 7.1 DNG converter, and I suppose that is what is used when I download from camera memory card through Bridge 5.1 and click dng conversion.
Same thing if I open the .nef in Photoshop 5.1 , which kicks up CameraRaw converter 6.7.0.339.
Why is this?
Can't .dng have more than 10-bit depth?
Sverk

Well, according to the user manual and to the review in
http://www.imaging-resource.com/PRODS/D90/D90A.HTM
the D90 delivers 12-bit color depth in the .NEF files.
Of course, I haven't looked at the actual pixel data to find out how finely graded they are.
What I'm looking at is what Bridge 5.1 (WindowsXP) says about the files in the
Metadata/ Bit depth entry.
In that, the .NEF files are listed as "16-bit" depth (although it will actually hold only 12-bit resolution), but when converted to .DNG it says only "10-bit",
and that holds both when the conversion is done automatically during the importing from the camera, and when converting from .nef files afterwards.
Archiving pictures in the .dng format seems to be a good idea -- but only if no information is lost in the conversion.
Thus, the "10-bit" info showing in Bridge worries me.
Might it be that the meaning of bit depth is different in the two file formats?
Might there be something about the de-mosaicing that necessarily consumes two bits of depth? Whether in the .dng conversion -- or when saved .nef files are later to be used?
In other words, for practical purposes, are the formats equivalent in color resolution,
Or is there indeed a certain loss?
Maybe a very difficult question, but I'd sure want to have a technical ly definite answer before I dare switch to using the .DNG format all the way.
Sverk

Turning on Render at Maximum Bit Depth and Maximum Render Quality crashes render every time

I've tried a few times to render an H264 version of my Red media project with Maximum Bit Depth and Maximum Render Quality. Premiere crashes every time. I have GPUs enabled. Are people using these settings with Red media and successfully rendering?

To answer your specific question did you see the tooltip?
I beleive it allows for 32-bit processing (16-bit if unchecked). Per the project settings help file at http://helpx.adobe.com/premiere-elements/using/project-settings-presets.html
Maximum Bit Depth
Allows Premiere Elements to use up to 32‑bit processing, even if the project uses a lower bit depth. Selecting this option increases precision but decreases performance.
The help file for export is somewhat less informative about what it actually does but does point out that it is the color bit depth - http://helpx.adobe.com/media-encoder/using/encode-export-video-audio.html
(Optional) Select Use Maximum Render Quality or Render At Maximum Bit Depth. Note: Rendering at a higher color bit depth requires more RAM and slows rendering substantially.
In practice the simplest suggestion is to export twice - once with / once without the setting and compare the time taken and perceived quality.
Cheers,
Neale
Insanity is hereditary, you get it from your children
If this post or another user's post resolves the original issue, please mark the posts as correct and/or helpful accordingly. This helps other users with similar trouble get answers to their questions quicker. Thanks.

Bit Depth Setting?

I feel like I'm missing something that's very obvious-- sorry if I am. How do I set the bit depth for a new audio file? I want to create 24 bit audio files, but as soon as I create a new file, STP automatically makes it 16bit. The only setting I've seen so far is in Preferences, to export at 24. I've set that, but nowhere does there seem to be any setting for recording or creating audio at 24bits.
Thanks for any help, and thanks for putting up with what's probably a dumb newbie question!
Mac OS X (10.4.7)

I believe all the current macs support 24 bit. You just need to set it in the project settings as mentioned above (assuming your machine supports it) and you're limited to recording in the project window. Recording in the waveform window seems to always record at 16 regardless of settings.

Choosing bit depth while recording audio

Hi,
First, the short version of the question:
Do you know of any simple audio recording app (ie: not a monster audio suite) which takes low CPU usage, and lets you choose the internal bit depth using for audio sampling (ie: sample at PCM16, at PCM24,...) ?
Now, in case you need it, the long version of the question:
I'm digitizing my huge collection of audio tapes, with my G5 iMac (OSX 10.4). The quality of the recordings is somewhat low, so I don't need a high-end audio card, and I believe the builtin audio card of this G5 is more than adequate for this task (Burr Brown PCM3052 with support for PCM16, PCM24, and AC3 16).
I've done my first tests with the just-released new version of Audacity (2.0.0), which is compatible with 10.4 and higher, so I'm very happy to be able to install a brand-new application in 10.4.
Audacity works fine, but has a couple of issues:
First, you can choose the audio track bit depth, but not the internal bit depth at which you're sampling. In other words, I'd expect to use some selector to choose from PCM16 or PCM24 recording, which is what this iMac seems to support. However, I can only choose the storage bit depth, not the sampling bit depth. I suppose Audacity is recording at PCM24, but I've no way to test this.
Second, it takes about 40% CPU usage while recording (yes, I disabled the realtime graphics update of Audacity... if I don't disable it, usage becomes 60% or higher).
I tried Audio Hijack Pro (well, the last version supporting 10.4), and it won't let me choose the sampling depth either.
So... do you know of any simple recording app which overcomes this issues?
TIA!
cesarpixel

although I was looking for something simpler.
Compared to what you describe, my use of GarageBand sounds "simpler."
GarageBand is part of iLife, so it is meant to be easy to use, unlike big brother Logic. Furthermore, I am using it in a very limited way. When recording, I just set up ONE track, which is pretty easy.
GB records at the best possible quality, up to 24-bit. I used 16-bit because my source was analog tapes and records, and I did not think it would make the final result any better. The GB project file serves as my "master copy" for future enhancement.
The best part is editing the full recording of the entire tape/record, into individual songs, then exporting them to iTunes as 256 kbps AAC (or MP3) files, all within GarageBand. A few of my records had skips, and I was able to edit them out by overlapping two tracks with the same portion of music, and carefully eliminating the skip at the transition between the two tracks. Also, because the source is analog, there is noticeable background hiss when the music level is low. It is particularly noticeable at the end of songs that fade out. To make the hiss fade out along with the music, I gradually adjusted the master volume level downward to zero starting when the music level became low enough for the hiss to be noticeable. That made the hiss fade out, which helped "hide" it.

Can an item of 1 Bit depth have layers?

Can an item of 1 Bit depth have layers which can be copied or duplicated. My 1 Bit item is partially locked and the layer menu does not permit any operations.

jaysachs wrote:
Does a Black/white item have a 1 Bit depth?
Yes. One bit binary is either on/off = black/white.
Cheers,
Neale
Insanity is hereditary, you get it from your children
If this post or another user's post resolves the original issue, please mark the posts as correct and/or helpful accordingly. This helps other users with similar trouble get answers to their questions quicker. Thanks.

Supported bit depth and sampling frequency

Hi,
Is it possible to determine at runtime what sampling frequencies and bit depths are available for audio capture -- other than by trial and error? It does not seem that way from the API, but maybe I missed something, so I decided to post a question here. The only thing I was able to find so far is that 8-bit PCM WAV is required to be supported by all devices that support audio at all.
Thanks in advance for any info!
Sergei

Thanks for the info. I guess I was looking for something more specific - the exact bitrates and sample rates that Creative claims to support. Would you know where official and comprehensi've data can be had? There must be a tech spec somewhere.
It is common these days in business to see a recording of, say, a conference call or seminar presentation at 32k bitrate/025Hz, or even 24k bitrate/8000Hz, posted to a company's website for download by those who could not be there, and MP3 players are increasingly used for their replay. Companies use low digitization rates because there is no need for hifi and the files are much smaller: less storage, faster download.
I'd be surprized to think that Creative don't have compatibility with the standard range of rates offered by ubiquitous programs like Audacity and dBpower, the latter being one they themselves recommend!

Apple Pro Res 422 HQ export - "render at max depth" and "24 bit or 48 bit depth"?

I'm exporting my 90 minute feature for DCP using the Apple Pro Res 422 HQ codec. The film is 1920x1080 and that is what the export will be. We used a variety of cameras (Canon 7D, Sony XR160, GoPro, Blackmagic in HD) for the film.
For the export options:
Do I check "Render at Maximum Depth"?
Which do I choose - 24 bit or 48 bit depth? - one has to be chosen even when "Render at Maximum Depth" is unchecked
When I asked the DCP house, they said that "Render at Maximum Depth doesn't actually do anything when using this codec" and haven't answered the 24 vs. 48 bit question.
This discussion:
https://forums.adobe.com/message/4529886#4529886
says that you "never need to enable the Max Render Quality (MRQ) unless you are exporting in a format/pixel ratio different from your original video."
This discussion:
https://forums.adobe.com/message/5619144#5619144
adds insight into what 24 vs 48 bit depth means, but doesn't answer my specific question
Thanks for your help.

For your reading enjoyment -
http://forums.adobe.com/message/4529886
http://images.apple.com/finalcutpro/docs/Apple_ProRes_White_Paper_October_2012.pdf
A question for you - what is your workflow where you think you might need this? Keep in mind that the majority of cameras only record 8-bit color, and also in a very highly compressed format, so you won't necessarily gain anything by going to 4444, as the source video is of limited quality already.
Basically, if you don't have any high-bit-depth sources in your timeline to preserve the quality of, there may be little or no benefit to enabling "Max Depth".
Thanks
Jeff Pulera
Safe Harbor Computers

Creative Audigy 2 NX Bit Depth / Sample Rate Prob

This is my first post to this form
Down to business: I recently purchased a Creative Audigy 2 NX sound card. I am using it on my laptop (an HP Pavilion zd 7000, which has plenty of power to support the card.) I installed it according to the instructions on the manual, but I have been having some problems with it. I can't seem to set the bit depth and sample rate settings to their proper values.
The maximum bit depth available from the drop down menu in "Device Control" -> "PCI/USB" tab is 6 bits and the maximum sample rate is 48kHz. I have tried repairing and reinstalling the drivers several times, but it still wont work. The card is connected to my laptop via USB 2.0.
I looked around in the forms and found out that at least one other person has had the same problem but no solution was posted. If anyone knows of a way to resolve this issue I would appreciate the input!
Here are my system specs:
HP Pavilion zd 7000
Intel Pentium 4 3.06 GHz
GB Ram
Windows XP Prof. SP 2
Thnx.
-cmsleimanMessage Edited by cmsleiman on -27-2004 09:38 PM

Well, I am new to high-end sound cards, and I may be misinterpreting the terminology, but the sound card is supposed to be a 24bit/96kHz card.
I am under the impression that one should be able to set the output quality of the card to 24bits of depth and a 96kHz sample rate, despite the speaker setting that one may be using, to decode good quality audio streams (say an audio cd or the dolby digital audio of a dvd movie.) I can currently achieve this only on 2. speaker systems (or when i set the speaker setting of the card to 2.) Otherwise the maximum bit depth/sample rate I can set the card output to is a sample rate of 48kHz and a bit depth of 6bits.
Am I mistaken in thinking that if I am playing a good quality audio stream I should be able to raise the output quality of the card to that which it is advertised and claims to have?
Thnx

Can't import a JPEG: "video bit depth of this file is unsupported"

Re above error message when attempting to import into a Premiere CS6 project. I used Media Info to inspect the file's properties as per below. All of the video in my project is 8bit video. Any help is much appreciated..
Image
Format                                     : JPEG
Width                                        : 1 306 pixels
Height                                       : 979 pixels
Color space                             : YUV
Bit depth                                   : 8 bits
Compression mode                : Lossy
Stream size                              : 2.83 MiB (100%)

Hi Jeff, thanks!!
I guess one cannot import cannot import a CMYK image in Premiere at all.
When I inspected the image in Media Info it ID'd it as a YUV and so I assumed that the imagwe was converted for use in a video that they'd created sometime in the past.
Is there even such a format as YUV for still images? Probably not even though Media Info ID'd it as such.
Anyway, Photoshop ID'd it as CMYK and when converted to RGB that did the trick!
Thanks!

Bit depth question

Similar Messages

Maybe you are looking for