# Linux hp Z book 10 bits

**URL:** <https://psychtoolbox.discourse.group/t/linux-hp-z-book-10-bits/110>\
**Category:** Uncategorized\
**Created:** [April 6, 2017, 12:56am UTC](https://psychtoolbox.discourse.group/t/linux-hp-z-book-10-bits/110 "2017-04-06T00:56:43Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![Allen\_Ingling](https://avatars.discourse-cdn.com/v4/letter/a/a183cd/32.png) [@Allen\_Ingling](https://psychtoolbox.discourse.group/u/Allen_Ingling)\
**Post date:** [April 6, 2017, 12:56am UTC](https://psychtoolbox.discourse.group/t/linux-hp-z-book-10-bits/110/1 "2017-04-06T00:56:43Z")

</div>

Dear Mario  
A preliminary report. I ran my test program on my Linux hp. It shows 10 but luminance precision. I can't send the graph yet because I don't know how to do email in Linux.   
  
What is the magic number for dither on this computer? I don't know the exact model but it's a Z Book. Same computer that hornet wrote to you about.   
  
I was surprised by a limitation. On macOS I enable Clut mapping with 4096 Clut size. Works fine. In Linux if the requested Clut size is larger than 256 the call to loadnormalizedgammatable with load=2 gives a fatal error complaining that my Clut is bigger than 256. Seems weird since it was already told when I enabled that I'd be using a 256 element soft Clut.   
  
Best  
Denis  
  
Sent from my iPhone

---

<div class="post-metadata">

**Author:** ![Allen\_Ingling](https://avatars.discourse-cdn.com/v4/letter/a/a183cd/32.png) [@Allen\_Ingling](https://psychtoolbox.discourse.group/u/Allen_Ingling)\
**Post date:** [April 6, 2017, 11:25pm UTC](https://psychtoolbox.discourse.group/t/linux-hp-z-book-10-bits/110/2 "2017-04-06T23:25:44Z")

</div>

On 04/06/2017 02:56 AM, Denis Pelli wrote:  

> \> Dear Mario A preliminary report. I ran my test program on my Linux  
> \> hp. It shows 10 but luminance precision. I can't send the graph yet  
> \> because I don't know how to do email in Linux.  
> \>  
> \> What is the magic number for dither on this computer? I don't know  
> \> the exact model but it's a Z Book. Same computer that hornet wrote to  
> \> you about.  
> \>  
> \> I was surprised by a limitation. On macOS I enable Clut mapping with  
> \> 4096 Clut size. Works fine. In Linux if the requested Clut size is  
> \> larger than 256 the call to loadnormalizedgammatable with load=2  
> \> gives a fatal error complaining that my Clut is bigger than 256.  
> \> Seems weird since it was already told when I enabled that I'd be  
> \> using a 256 element soft Clut.  
> \>

I don't understand that? What kind of clut mapping with load=2? On Linux   
the driver uses the discrete 256 slot hardware gamma table, instead of   
the non-linear gamma mapping that macOS now uses. Also PTB on Linux   
completely disables hw gamma tables in \>= 10 bit modes, so all gamma   
correction is done via PsychColorCorrection(). You start off with a   
identity gamma table.  
  

> \> Best Denis  
> \>  
> \> Sent from my iPhone  
> \>