MItAC update

- rough notes on vinyl physics available here:
http://www.music.mcgill.ca/~deleon/phonograph_notes.doc
- above document will be updated regularly as readings continue
- main ideas from AES articles highlight imperfections of vinyl playback that inevitably lead to upper harmonic distortions, perhaps these can be avoided optically and then physically modeled back into the clean audio for a comparison study of preference

MItAC update

- reading and note-taking continuing on vinyl physics, a few interesting facts that may have direct consequence to our study

• cutters use carriage that moves radially across vinyl, whereas playback turntables typically rotate from a pivot. Thus such turntables suffer from unavoidable pitch/azimuth errors
• RIAA reproduction EQ curve (the inverse of the RIAA lathe cutter curve) does not account for the typical 50kHz rolloff in most lathes. Optical methods could allow properly compensating reproduction curve, as well as compensate for non-standard EQ curves used prior to standardization
• curvature of playback stylus greatly determines amplitude of certain frequencies i.e. elliptical stylus fits into tightly packed modulations better and allows for more displacement, thus explaining better response to higher frequency. Optical method completely avoids this and is now only limited by groove space requirements and capabilities of cutter
• tracing distortion (caused by differences in linear travel speed of stylus across record width) is unavoidable, and results in different frequency performance in inner and outer grooves. Outer modulations perform better at higher frequencies since modulations are stretched out, which is the same principle as improving sound quality via increasing rpm. Again optical method completely avoids this in theory

- updated Wiki to include vinyl physics links found on the internet. Draft status
- EndNote database started to incorporate large body of work in Journal of AES on phonographs

MItAC update

- new software arriving to resolve Vision stitching problems
- all future bugs and software issues will be posted at http://visionbugs.blogspot.com/ in addition to our personal blogs for organization and to avoid redundancy
- wealth of phonograph studies on JAES, currently reading and note taking
- current plan is to have a one-on-one tutorial Friday afternoon to teach each other what we have learned thus far (Damon=interferometry, Simon=vinyl physics)
- microscope software familiarity work will halt until new software arrives

Vision stitching bugs

1) Circular/Annular stitching does not support stitch from current location. Specifying an inner / outer diameter with "Use Current Location" starting option selected will still result in a modal dialogue that disables other actions when taking a new measurement. The modal dialogue requests for the user to specify boundaries for circular stitching, which can only be completed in the stitching "Teach" procedure.

2) Crashing Vision in Circular/Annular stitching mode
a) We specify inner/outer dimensions manually with Turret = 5x , FOV = 1x

• Outer left edge of the sample: in our test, left corner of the "Y"
  
• Outer top edge: two grids up and right from the origin
  
• Outer bottom edge: 4 grids down from upper top edge point
  
• Inner bottom edge: 1 grid up from bottom edge point
• Resulting diameters
• Outer: 0.4765 mm
• Inner: 0.2315 mm
  

b) Select "New Measurement" results in a modal dialogue requesting the manual specification of outer left edge followed by clicking "Next". The "Next" button is greyed out and only clicking "Finish" will start stitching.

c) During measurement, the microscope appears to attempt to scan at a larger area than specified. It starts correctly at the outer left edge, but the other locations chosen seem to be offset from the other locations specified (outer top edge, outer bottom edge, inner bottom edge). From our limited observations, the three boundaries seem to be moved up and to the right from their originally specified locations.

d) After taking measurements and camera returns to outer left edge, Vision crashes with no prompt or error message.

3) Freezing and crashing Vision when attempting to use Stage File stitching using a manually created stage file from X/Y Grid. See Damon's blog (http://damonli.blogspot.com/) for complete description. Will contact Veeco sales tomorrow with regards to bugs encountered for software update. We are currently using 3.50, released 8/8/05.

4) Familiarization with Wiki syntax and summaries of VisualAudio and LBNL. Hierarchical structure of http://coltrane.music.mcgill.ca:16080/DDMAL/index.php/Preservation to be decided with Damon tomorrow.

MItAC update (July 19, 2006)

- obtained reading material for phonograph/gramophone characteristics
- rectangular stitching familiarization: specify length/width, and current FOV will be lower left
- circular stitching familiarization: specify inner/outer diameter (circular or annulus), optionally can select "learn" to manually pinpoint outer left edge/top edge/bot edge, etc., where it seems we were sometimes able to crash Vision, and once crash Windows in this manual mode (suspected it may be due to improper manual specification of annulus)

- blue screen: "The problem seems to be caused by the following file: bcam.sys"

- PAGE_FAULT_IN_NONPAGED_AREA
- Technical Information: STOP: 0x00000050 (0xE0617000, 0x00000001, 0xF77192DD, 0x00000000)
- bcam.sys

- Address F77192DD
- Base at F7712000
- Date stamp 416ffae0

- stage file stitching familiarization: not yet started
- ignoring bug, it appears that annulus stitch could be used to scan an entire record, so long as backscan and length are sufficient to account for surface height changes

interesting links for vinyl physics

-  http://www.vinylrecorder.com/stereo.html illustrates stereo groove modulation.  While vertical and lateral modulation were both used for 78rpm mono, lateral modulation became the standard because of groove durability and the fact that large amplitudes of vertical modulation would virtually eliminate the groove.  Instead of implementing stereo with vertical/lateral, problem of vertical avoided by compromising both sidewalls to modulate at 45deg, phase-shifted 180.
-  http://en.wikipedia.org/wiki/Gramophone_record wikipedia article, great overall picture from history to materials and manufacturing
-  http://www.aardvarkmastering.com/riaa.htm RIAA dimensions and standards
-  http://www.euronet.nl/~mgw/background/riaa/uk_riaa_background_1.html RIAA equalization graphs
-  http://eil.com/explore/guide/vinyl_making.asp great background info on cutting vinyl master discs to manufacturing stampers
-  http://www.vestax.com/v/products/recorders/vrx2000.html (Vinyl lathe, ~$10,000 USD), http://www.vinylrecorder.com/order.html (Vinyl lathe, ~$4000 euro)
-  http://www.music.mcgill.ca/~deleon/courses_files/mumt611_files/mumt611_a6_deleons.pdf my presentation on image to audio conversion for MUMT611, an overview of its current state as of April 2006
-  http://www-cdf.lbl.gov/%7Eav/ LBNL sound reproduction page containing movies, audio, and papers of Fadeyev and Haber's work for wax cylinders

MitAC update

-  Naming convention for images agreed upon 33/45/78-backscan-length-modthresh-datarestorepixels-fov-turret-Z
-  Choice of known experimental 78rpm for intial studies - "There's a tear in my beer tonight" by Hoosier Hot Shots and Sally Foster, Damon is searching for clean digital copy and we plan to record it from a turntable
-  Platter, record mat and approximate thread match found, longer screws and washplates to be purchased Tuesday from Home Depot to complete microscope stage assembly
-  Partial delegation of short-term research responsibilites: Damon (physics of white light interferometry, stitching methods), Simon (physics of vinyl manufacturing/cartridges/lathes, Wikipedia literature update for LBNL etc., complete platter stage assembly), both (familiarization with imaging capabilities to the level of Tom's images and to begin stitching feasibility study)
-  Currently capable of achieving images as described by Veeco salesman in tutorial (clear top/bottom groove, no sidewalls), this is partially circumvented by enabling data restore (linear interpolation) but is not of the quality of Tom's VSI images in Ich's powerpoint slides
-  the problem appears to be focusing in deep cracks (~30um) rather than post processing since fringe patterns do not appear on sidewalls regardless of tilt, filter, fov, turret, intensity, and fringes appear successfully on top groove and bottom during scan
-  ASCII datasets seem to be the choice output format due to options to incorporate post-processing and better than 8-bit resolution, rather than TIFF as previously suggested