# SVD note review (lepton ID)
### Page1
> We improve the identification performance of low-momentum leptons, especially electrons, using the specific ionisation information from the silicon-strip vertex detector (SVD) of the Belle II experiment.
- [ ] We improve the identification performance of low-momentum leptons, especially electrons, by including the specific ionisation information from the silicon-strip vertex detector (SVD) into the Belle2 PID framework.

> The study of electrons is based on a 6.2fb−1 privately generated MC and proc11 data from hlt hadron skim
- [ ] The calibration and performance study of the electrons are performed with proc11 dataset and privately generated Monte Carlo.
- [x] Q: from hlt hadron skim -> you used hlt_hadron skim for electron PID studies ?

> We also study the resolution of SVD to separate muon and pion dE/dx distributions using tracks generated by Particle gun module. The study was repeated for muon tracks from two-photon process e+e− → e+e−μ+μ− and slow pion tracks from D∗ → D[Kπ]π control sample
- [ ] The muons calibration and performance is considered as similar to pions due to similar mass. The assumption has been verified by studying their dE/dx bands (mean/resolution) with particle gun based MC. This is further verified in data(?) by studying muons from two-photon process e+e− → e+e−μ+μ− and slow pion tracks from D∗ → D[Kπ]π decays sample.

### Page3
> Combining information from a number of subdetectors we identify the final-state particles to study the underlying physics processes happening inside the detector. At Belle II [1] we use a likelihood based PID algorithm to add information from various subdetectors towards identifying a track as a lepton [2]
- [ ] The information from a number of subdetectors is combined to identify the final-state particles, those are then used to study the underlying physics processes happening inside the detector. At Belle II [1] we use a likelihood based PID algorithm to add information from various subdetectors towards identifying a charged track (e.g. lepton or hadron) [2]

> This algorithm heavily relies on the information provided by the electromagnetic calorimeter (ECL) for identifying electrons and the KL0 and muon detector (KLM) for identifying the muons A track must have a minimum momentum of 0.3 and 0.7GeV/c to be able to reach the ECL and KLM, respectively.The threshold results in a poor performance of the algorithm for low-momentum tracks that fail to reach these subdetectors. Using the specific ionisation information from the silicon-strip vertex detector (SVD) of the experiment, we aim to improve the lepton identification performance for such tracks.

- [ ] For leptons, this algorithm heavily relies on the information provided by the electromagnetic calorimeter (ECL) for identifying electrons and the KL0 and muon detector (KLM) for identifying the muons. A track momentum threshold of 0.3 GeV/c and 0.7GeV/c is required to reach ECL and KLM detectors. This requirement, therefore, limits the performance of lepton ID below the momentum threshold. The specific ionisation information from CDC (Charge Drift Chamber) and SVD plays an important roles as they are capable of providing PID in such low momentum region. Herewith silicon-strip vertex detector (SVD), our aim is to improve the lepton identification for such low-p tracks and add PID constraint to higher momentum region.

> Figure a/b
- [ ] If these figures are prepared from unofficial MC so I am not sure if you can write BELLEII on plots (please double check).
- [ ] put axis label in same alignment style
- [ ] are you using BestCandidate selections here ?

> 2. SVD CALIBRATION
- [ ] 2. SVD Electron calibration

> Generic MC: 6.2 fb−1 equivalent of privately generated MC data. 3.1 fb−1 for calibra- tion and performance study each.
- [ ] Generic MC: We used 3.1 fb−1 equivalent of privately generated MC for calibration and same size (3.1 fb-) for the performance study.
> Data: proc11
- [ ] Proc11 Data:

### Page 4
> 2.2 Reconstruction
- [ ] 2.2. Track selections?

> interaction point and instead require them
- [ ] interaction point but we require them

> tracks satisfying these selection criteria
- [ ] tracks satisfying SVD hit criteria

> backgrounds such as (radiative) Bhabha
- [ ] backgrounds such as radiative-bhabha,

> To purify the sample,
- [ ] To purify the sample further,

> Figure4
Why you split this figure (also in the appendix) into a set of two

> by decreasing the range of Me+e− to 10 MeV/c2
- [ ] you are using 3-10 range only? so what do you mean by decreasing here?

> The plots for the invariant mass distribution with various cuts are shown..
- [ ] Please add extra explanation(various? or with and without ?). At this point, please also add a table to summarize all the cuts (for e- and photon) you have used to improve sample quality.

### Page 5
> Figure 3 caption
- [ ] same above comment: prepare a summary table of all cuts and cite that table in the caption here.

> (Apart from the gain in signal purity, reducing the range of Me+e− to 10 MeV/c2 makes it easier to model the background with a first order polynomial)
- [ ] remove or move these texts to the above section.

### Page8
- [ ] Please compare Data vs MC performance as well to see how much they agree/dis-agree.

### Page9
> SVD DE/DX RESOLUTION FOR MUONS
- [ ] SVD Muons Calibraiton

> vs. momentum as shown in Fig 13.
- [ ] vs. momentum as shown in Fig 13b.

### Page11
> Add more text about decay mode in the figure caption

## General comments
- [ ] Fix figure in chronological order
- [ ] Proc11 official is processed without SVD so I am not sure how you verified your electron performance with proc11-exp10 data. Don't you need to reprocess proc11-exp10 again with a new electron payload?
- [ ] add lines numbers
- [ ] plans for future calibration?
- [ ] Do you have performance plots seperated by charge (skip if complicated)?
- [ ] PID performance as a function of costh (new on May 9)?